I have an older instance of Telegraf and Grafana running and I am simply attempting to rebuild it. The Grafana and Telegraf configurations are exactly the same but there are about 4 instances of inputs.snmp that seem to show NoData every 5 minutes and then resolve after 1 minute. These are exact times.On the previous version I don’t have this issue.
Increasing the Interval seems to resolve this issue but it also changes the data collected ofcourse. Is there any way I can optimise this further? I am assuming it is Telegraf being slow to collect data based off of this but not 100% sure.
Try running Telegraf with a second output to a file and see if the data that is missing from Grafana exists in the output file or not. I have run into this same situation more than once and there are so many possibilities but first lets see if Telegraf is even collecting the data. Second run Telegraf with –debug and see if you see any logs for metric drops or buffer limits. You could also check your grafana interval and see what it is set to.
Thanks for the feedback! Not sure if this is the best way to fix it but I’ve added ‘createEmpty = false’ into my Flux query. This still looks to error when there actually is no data but not when Grafana falsely believes there is no data.
